MATLAB is a popular programming language in many engineering disciplines, intended for the fast development of prototypes. But as we all know, prototypes make it into production all the time, so now you're stuck. Unfortunately, there are no style checkers or "good" static analysis tools for MATLAB. This project attempts to fill this gap.
If you have MATLAB (or MATLAB embedded in Simulink models) in your production code and want to improve code quality then this tool-suite is for you.
MISS_HIT comes with the following tools, all of which come with user manuals and setup instructions: https://florianschanda.github.io/miss_hit/
-
Style Checker
mh_style
A simple coding style checker and code formatter for MATLAB or Octave code, including MATLAB embedded inside Simulink models.
-
Code Metrics
mh_metric
A simple code metric tool for MATLAB or Octave code, including MATLAB embedded inside Simulink models.
-
Bug finding
mh_lint
A simple linter for MATLAB or Octave code, including MATLAB embedded inside Simulink models.
-
Code/Test traceability
mh_trace
A simple tool to extrace tags from test and code to demonstrate requirements traceability.
-
Diff helper
mh_diff
A tool for diffing MATLAB code inside Simulink models. Note that other changes (e.g. different connections) are not detected; this is only working for embedded MATLAB.
-
Copyright notice helper
mh_copyright
A tool that can update or adjust copyright notices in bulk. Helpful if your company changes name, or you have year ranges that need updating.
Please refer to the release notes for a summary of recent changes and known issues.
We intend to provide more tools later, please refer to the roadmap for more information.
$ pip3 install --user miss_hit
This installation also adds five executable scripts mh_style
,
mh_metric
, mh_lint
, mh_copyright
, and mh_diff
into
.local/bin
, so please make sure that this is on your PATH
.
You can also use the python -m
syntax to directly invoke the
program. This might be useful if you're on a heavily locked-down
corporate Windows environment:
$ python3 -m miss_hit.mh_style
To use MISS_HIT you just give it a set of files or directories to process, for example:
$ mh_style my_file.m
$ mh_style --process-slx my_model.slx
$ mh_style src/
Configuration and setup is described in the user manuals
It is recommended to use pip, as that gets you the latest stable release. However, it is possible to directly use MISS_HIT from a checkout. MISS_HIT does not require any python packages or libraries. Just check out the repository and put it on your path. That's it.
The version of Python I am using is 3.6
but any later version should
also work. I am not using any overly fancy language features.
If you want to help develop you will need Linux as the test-suite doesn't really work on Windows. You will also need Pylint, PyCodeStyle, Coverage, and Graphviz. Install as follows:
$ apt-get install graphviz
$ pip3 install --user --upgrade pylint pycodestyle coverage
For publishing releases (to GitHub and PyPI) you will also need:
$ pip3 install --user --upgrade setuptools wheel requests
There are serious issues present in the MATLAB and Octave languages on
all levels (lexical structure, parsing, and semantics) that make it
very difficult to create any tool processing them. In fact, GitHub is
littered with incomplete attempts and buggy parsers. The usual
question is "but what about Octave?"; it is a similar language, but it
is not compatible with MATLAB. If your problem is parsing MATLAB then
the Octave parser will not help you. Even very simple statements such
as x = [1++2]
mean different things (3
in MATLAB, syntax error in
Octave).
I have documented the key issues we've faced and how we've resolved them.
The basic framework, style checker and code metrics tool of MISS_HIT
(everything under miss_hit_core
) are licensed under the GNU GPL
version 3 (or later) as described in
LICENSE.
The advanced analysis tools of MISS_HIT (everything under miss_hit
)
are licensed under the GNU Affero GPL version 3 (or later) as
described in
LICENSE.AGPL.
The vast majority of this work is (C) Florian Schanda. Contributions from the following people and entities are under their copyright, with the same license:
- Alina Boboc (Documentation style)
- Benedikt Schmid (MATLAB integration)
- Remi Gau (CI/CD templates)
- Veoneer System Software GmbH (JSON Metrics)
- Zenuity AB (Key parts of the lexer)
This project includes modified/adapted parts of the GNU Octave
testsuite under tests/parser/octave_*
. These are (c) their original
authors. Each file there describes from which file they derive.
Some of the parser tests include code samples and documentation snippets from the publicly available MathWorks website. An attribution (in comment form) is always included in these cases.
The documentation uses feather icons which are licensed under the MIT License.